Prompt delivery of cardiopulmonary resuscitation (CPR) increases survival from sudden cardiac arrest (SCA) by 2-3 fold, yet most bystander-witnessed victims still die without an initial CPR attempt. There is a critical knowledge gap in our understanding of how to disseminate CPR training to at-risk populations. This is important as CPR delivery represents one of very few tools available to improve survival from SCA, which afflicts over 200,000 people in the United States each year. Individuals with coronary disease are at significantly increased risk of SCA, and greater than 75% of SCA events occur in the home environment. Taken together, these two observations highlight the opportunity to impact cardiovascular health by targeting CPR training to family members (and especially spouses) of patients hospitalized for complications of coronary disease. Current CPR training implementation largely fails to reach this at-risk demographic. Attempts to simplify CPR education and thereby broaden and increase training have led to the development of hands- only CPR, omitting the need for mouth-to-mouth ventilation. However, the preferred strategy for broad implementation is unknown. The long term goal of our work is to establish best practices for real world CPR dissemination that match training venues with at-risk populations, maximize resuscitation skill retention, and promote willingness to act. Towards this end, the goal of this application is to test two CPR educational strategies using a recently validated video-self instruction (VSI) CPR program in a hospital-based implementation program. Our central hypothesis is that hands-only CPR training will yield improved skills retention and address important training barriers when compared to standard CPR in this highly motivated at- risk population. As a dissemination project, a high degree of autonomy will be given to local stakeholders at each hospital to tailor the staff and enrollment approach at each site, and an important objective will be to learn from their implementation experiences. To test our central hypothesis, we will pursue the following Specific Aims in an eight hospital cohort: (1) Assess objective metrics of resuscitation skill retention at three months in subjects randomized to hands-only VSI CPR training versus standard VSI CPR. (2) Measure the perceived barriers and motivations to enrollment for CPR training in our hospital-based model using mixed quantitative and qualitative survey methods, and (3) Analyze the costs of our VSI CPR educational program, and perform a budget impact analysis comparing the two CPR strategies in Specific Aim 1 with regard to the staff opportunity costs and cost per subject with adequately retained skills as cost modeling inputs. Our investigative team is uniquely poised to accomplish this work, with its longstanding focus on clinical CPR research and track record of collaboration in CPR education. This innovative approach, to match CPR training with public need via the hospital platform and the use of VSI, may serve as a national model for dissemination of lifesaving CPR skills to a highly motivated population.